Research
Conference Proceedings
Learning in Routing Games with Symmetric Incomplete Information (joint with Tristan Tomala and Marco Scarsini) was accepted and published (abstract) in WINE2020. My talk is accessible here. An extended version has been submitted to a journal.
Journal Articles
Bayesian Learning in Dynamic Nonatomic Routing Games (joint with Tristan Tomala and Marco Scarsini). Published in Games and Economic Behavior (2023).
Abstract: We consider a discrete-time nonatomic routing game with variable demand and uncertain costs. Given a routing network with single origin and destination, the cost function of each edge depends on some uncertain persistent state parameter. At every period, a random traffic demand is routed through the network according to a Wardrop equilibrium. The realized costs are publicly observed and the public Bayesian belief about the state parameter is updated. We say that there is strong learning when beliefs converge to the truth and weak learning when the equilibrium flow converges to the complete-information flow. We characterize the networks for which learning occurs. We prove that these networks have a series-parallel structure and provide a counterexample to show that learning may fail in non-series-parallel networks.
Working Papers and Current Projects
Stochastic Consensus and the Shadow of Doubt (Job Market Paper)
Abstract: We propose a stochastic model of opinion exchange in networks. Consider a finite set of agents organized in a fixed network structure. There is a binary state of the world and, ex ante, each agent is informed either about the true state of the world with probability α or about the wrong state with probability 1-α. We model beliefs as urns where white balls represent the true state and black balls the wrong state. Communication happens in discrete time and, at each period, agents draw and display one ball from their urn with replacement. Then, they reinforce their urns by adding balls of the colors drawn by their neighbors. We show that this process converges almost-surely to a stable state where all urns have the same proportion of balls. We show that this limit proportion is a random variable with full support over [0,1]. We propose a conjecture on the distribution of this limit proportion based on simulations.
Cooperation in Stochastic Revision Games with Frequency-Dependent Payoffs
Abstract: We study the influence of time preferences on the existence of cooperative strategies in a model where two competing countries choose their level of emissions repeatedly. To do so, we build a model of revision games with flow payoffs and a cumulative state of the world to structure the trade-off faced by states involved in environmental transition. Two agents play a prisoners' dilemma over a finite time interval and are offered to revise their choice of action at stochastic dates. These dates are determined by the ticks of a common Poisson clock, hence the probability to act at ulterior dates decreases as time advances. At any point in time, the action profile yields a flow of payoffs and determines a common state of the world that captures the share of time each player has played the non-cooperative action. This state models the impact of countries emissions in the long-run. When time reaches the end of the interval, each player receives a terminal payoff that depends only on the state of the world. Our main interest is to study both the inuence of this decreasing discount factor on the optimal emission trajectories and the existence of equilibria sustaining cooperative strategies. In that respect, our work focuses first on proving the existence of Markov Perfect Equilibria, then showing that the existence of cooperative equilibria cannot be guaranteed by direct application of a Folk Theorem. We then focus on characterizing optimal threshold strategies and study how the timing of the game, the payoff structure and the parameter of the Poisson clock may jointly sustain cooperation.
Social Learning in Cooperative Games with Transferable Utility, with François Doyelle (LEMMA, Paris 2)
Weighted Pairwise Stable Network Formation in Incomplete Information, with Julien Fixary (CES, Paris 1)
An Axiomatic Approach to Biform Games, with Julien Fixary (CES, Paris 1)
Dynamic Agent Motivation with an Informed Principal, with Frédéric Loss (CY Cergy)
PhD Thesis
Information, Coordination and Cooperation: Essays on Learning in Dynamic Games, defended on Dec. 7th, 2021 at HEC Paris.